神经程序综合论文趋势 2023–2026

主要结论

• 神经程序综合论文从 2023 年的 57 篇增长到 2026 年的 59 篇，三年间扩大到约 1.0 倍。
• CVPR 是发神经程序综合论文最多的会议，2023–2026 年共 89 篇，占四个会议总量的 43%。
• 交叉方向里涨得最快的是驾驶仿真：同时属于两个主题的论文从 2023 年的 5 篇增至 2026 年的 10 篇。
• 在「程序综合」的 5 个主题里，它的 2026 年论文数排第 2；同组增长最快的定理证明是 13 倍，神经程序综合是 1.0 倍。

代表论文

按年份和会议均匀挑选，方便快速看到每一届的样貌

• NeurIPS 2026Feedforward Novel View Synthesis for Heterogeneous Cameras
• ICML 2026Search Space Synthesis for Parametric Functions
• ICLR 2026LLM as an Algorithmist: Enhancing Anomaly Detectors via Programmatic Synthesis
• CVPR 2026Beyond Geometry Artistic Disparity Synthesis for Immersive 2D-to-3D
• NeurIPS 2025AlgoTune: Can Language Models Speed Up General-Purpose Numerical Programs?
• ICML 2025High Dynamic Range Novel View Synthesis with Single Exposure
• ICLR 2025Execution-guided within-prompt search for programming-by-example
• CVPR 2025BimArt: A Unified Approach for the Synthesis of 3D Bimanual Interaction with Articulated Objects
• NeurIPS 2024HYSYNTH: Context-Free LLM Approximation for Guiding Program Synthesis
• ICML 2024Amortizing Pragmatic Program Synthesis with Rankings
• ICLR 2024Generating Pragmatic Examples to Train Neural Program Synthesizers
• CVPR 2024SuperSVG: Superpixel-based Scalable Vector Graphics Synthesis
• NeurIPS 2023Construction of Hierarchical Neural Architecture Search Spaces based on Context-free Grammars
• ICML 2023XAI Beyond Classification: Interpretable Neural Clustering
• ICLR 2023Language Models Can Teach Themselves to Program Better
• CVPR 2023Implicit Neural Head Synthesis via Controllable Local Deformation Fields
• NeurIPS 2026GPU-Accelerated Synthesis of Mixed-Boolean Arithmetic: Beyond Caching
• ICML 2026CHESS: Chebyshev Spectral Synthesis for Trajectory Condensation
• ICLR 2026RefineStat: Efficient Exploration for Probabilistic Program Synthesis
• CVPR 2026From Rays to Projections Better Inputs for Feed-Forward View Synthesis

全部神经程序综合论文

共 20 篇，第 1 / 1 页

1. NeurIPS 2025AlgoTune: Can Language Models Speed Up General-Purpose Numerical Programs?
2. NeurIPS 2025BIPNN: Learning to Solve Binary Integer Programming via Hypergraph Neural Networks
3. NeurIPS 2025Dynamic View Synthesis as an Inverse Problem
4. NeurIPS 2025EnCompass: Enhancing Agent Programming with Search Over Program Execution Paths
5. NeurIPS 2025GSRF: Complex-Valued 3D Gaussian Splatting for Efficient Radio-Frequency Data Synthesis
6. NeurIPS 2025HoliGS: Holistic Gaussian Splatting for Embodied View Synthesis
7. NeurIPS 2025HyPlaneHead: Rethinking Tri-plane-like Representations in Full-Head Image Synthesis
8. NeurIPS 2025Learning Neural Exposure Fields for View Synthesis
9. NeurIPS 2025LongMagpie: A Self-synthesis Method for Generating Large-scale Long-context Instructions
10. NeurIPS 2025Native-Resolution Image Synthesis
11. NeurIPS 2025Neural Attention Search
12. NeurIPS 2025Once Upon an Input: Reasoning via Per-Instance Program Synthesis
13. NeurIPS 2025PoE-World: Compositional World Modeling with Products of Programmatic Experts
14. NeurIPS 2025Point Cloud Synthesis Using Inner Product Transforms
15. NeurIPS 2025Program Synthesis via Test-Time Transduction
16. NeurIPS 2025RoomEditor: High-Fidelity Furniture Synthesis with Parameter-Sharing U-Net
17. NeurIPS 2025Searching Latent Program Spaces
18. NeurIPS 2025Single-pass Adaptive Image Tokenization for Minimum Program Search
19. NeurIPS 2025TF-MAS: Training-free Mamba2 Architecture Search
20. NeurIPS 2025Training Language Models to Generate Quality Code with Program Analysis Feedback

常见问题

2023–2026 年四大会议一共有多少篇神经程序综合论文？

共 208 篇，按年份分别是 2023 年 57 篇、2024 年 47 篇、2025 年 45 篇、2026 年 59 篇。

NeurIPS 2026 有多少篇神经程序综合论文？

NeurIPS 2026 有 18 篇神经程序综合论文，上一届是 20 篇。

ICLR 2026 有多少篇神经程序综合论文？

ICLR 2026 有 11 篇神经程序综合论文，上一届是 5 篇。

ICML 2026 有多少篇神经程序综合论文？

ICML 2026 有 10 篇神经程序综合论文，上一届是 2 篇。

CVPR 2026 有多少篇神经程序综合论文？

CVPR 2026 有 20 篇神经程序综合论文，上一届是 18 篇。

哪个会议的神经程序综合论文最多？

CVPR，2023–2026 年共 89 篇。